PROPOSAL SUMMARY Despite considerable investments in prevention and treatment of obesity in children, prevalence continues to increase with catastrophic individual and societal costs. Pacific Islanders are among the most at risk, with obesity prevalence rising disproportionately faster among Pacific Islanders compared to other ethnic groups. Obesity prevention strategies are clearly needed, but a) the early determinants of obesity among Pacific Islanders are poorly understood, and b) for optimal efficacy, it is necessary to first understand both the age at which elevated cardiometabolic risk markers are established and the periods of growth that are most sensitive to environmental or behavioral risk factors during childhood. As such, we established the Ola Tuputupua?e ?Growing Up? cohort in Samoa; the first longitudinal cohort study of children among the Pacific Island nations. To date, the cohort includes 450 children and their biological mothers with serial anthropometric measurements from 2015 (2-4 years old) and 2017-2018 (3.5-7 years old). Cross-sectional findings from the study demonstrate that this cohort is ideal for research focused on childhood obesity given the high prevalence and early emergence of cardiometabolic risk. The objective of this proposal is to develop a comprehensive model of childhood obesity development in Samoa that 1) describes growth and body composition during a critical age period for obesity development, 2) captures influences of individual and household factors and the age periods during which growth is most sensitive to them, and 3) quantifies the influence of longitudinal growth on cardiometabolic disease markers. We will focus on body mass index (BMI) to assess growth over time, diet and physical activity as modifiable individual-level factors, and household urbanicity and socioeconomic status as structural household- level factors. We plan to collect data in 2019 from the same children (5.5-9 years old) to provide a third time point for analysis, utilize newly collected dual energy-x-ray absorptiometry body composition data, and employ longitudinal modeling approaches to address the following specific aims: 1) Use multilevel modeling to identify individual and household-level factors associated with childhood BMI trajectories, 2) Use multilevel modeling to assess the association of childhood BMI trajectories with (a) body composition (b) blood pressure, and (c) glycosylated hemoglobin at 5.5-9 years old, and 3) Use structural equation modeling to assess the direct and indirect effect of individual-and household-level factors on childhood BMI and body composition. This research will be among the first to use a longitudinal, multilevel design to examine the time-fixed and time-varying effects of individual and household-level factors on growth, body composition, and cardiometabolic disease markers among children in the Pacific Island nations. The findings will likely apply to other settings and enhance our understanding of the range of bio-behavioral-environmental etiologic pathways of obesity onset and cardiometabolic risk development during childhood to concretely inform intervention strategies.